Official Pytorch Implementation of Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think (ICLR 2025)

Official PyTorch Implementation of "History-Guided Video Diffusion"

Fully open reproduction of DeepSeek-R1

A PyTorch library for implementing flow matching algorithms, featuring continuous and discrete flow matching implementations. It includes practical examples for both text and image modalities.

Unofficial implementation of Titans, SOTA memory for transformers, in Pytorch

CALVIN - A benchmark for Language-Conditioned Policy Learning for Long-Horizon Robot Manipulation Tasks

[RSS 2024] Code for "Multimodal Diffusion Transformer: Learning Versatile Behavior from Multimodal Goals" for CALVIN experiments with pre-trained weights

Implementation of "Conditional Score Guidance for Text-Driven Image-to-Image Translation" (NeurIPS 2023).

Memory-optimized training scripts for video models based on Diffusers

[NeurIPS 2024] CV-VAE: A Compatible Video VAE for Latent Generative Video Models

SEED-Voken: A Series of Powerful Visual Tokenizers

xLSTM as Generic Vision Backbone

Official repository of the xLSTM.

The first open autoregressive foundational video AI model.

VideoVAE+: Large Motion Video Autoencoding with Cross-modal Video VAE

VMC: Video Motion Customization using Temporal Attention Adaption for Text-to-Video Diffusion Models (CVPR 2024)

Autoregressive Model Beats Diffusion: 🦙 Llama for Scalable Image Generation

Official JAX implementation of xLSTM including fast and efficient training and inference code. 7B model available at https://huggingface.co/NX-AI/xLSTM-7b.

Official inference repo for FLUX.1 models

OmniGen: Unified Image Generation. https://arxiv.org/pdf/2409.11340

Tips for Writing a Research Paper using LaTeX

Adaptive Caching for Faster Video Generation with Diffusion Transformers

Offical codes for "AutoVFX: Physically Realistic Video Editing from Natural Language Instructions."

Inference script for Oasis 500M

The best OSS video generation models

Meta Lingua: a lean, efficient, and easy-to-hack codebase to research LLMs.

[NeurIPS 2024 D&B Track] Official Repo for "LVD-2M: A Long-take Video Dataset with Temporally Dense Captions"

Official repository for "iVideoGPT: Interactive VideoGPTs are Scalable World Models" (NeurIPS 2024), https://arxiv.org/abs/2405.15223